R Notebook

Import library

library(tidyverse)
library(ggplot2)
library(gridExtra)

# knitr::opts_chunk$set(fig.height=15) 

# library(usethis) 
# usethis::edit_r_environ()

Read file

# real file name
df = read.csv('../dataset/cleaned/cleaned_dataset.csv')


# just for testing

# df = read.csv('../dataset/cleaned/sample_cleaned_dataset.csv')

# print(names(df))

head(df, 5)

Completion time by assignment_gs_correct_count, assignment_gs_count, assignment_price, microtasks_count.

Each plot is categorized by assignment_status, device_category, assignment_type

prepare.df.by.category = function(cat){
  
  data.by.cat = df %>% select('assignment_gs_correct_count',  'assignment_gs_count', 'assignment_price', 'microtasks_count', 'completion.time.in.minutes', cat)
  
  data.by.cat = data.by.cat %>% 
  pivot_longer(cols=c('assignment_gs_correct_count',  'assignment_gs_count', 'assignment_price', 'microtasks_count'), names_to = 'parameter')
  
  return(data.by.cat)
}

df2 = prepare.df.by.category('assignment_status')

Note: Using an external vector in selections is ambiguous.
ℹ Use `all_of(cat)` instead of `cat` to silence this message.
ℹ See <https://tidyselect.r-lib.org/reference/faq-external-vector.html>.
This message is displayed once per session.

plt = df2 %>% ggplot(aes(x=value, y=completion.time.in.minutes, color=assignment_status)) +
  geom_point(alpha = 0.5) +
  facet_wrap(vars(parameter), nrow=1, scales = "free_x") +
  xlab('') + ylab('completion time (in minute)') + 
  ggtitle('plot by assignment status') + 
  theme(
    strip.text = element_text(size = 5),
    axis.title.y = element_text(size = 8),
    legend.position = 'bottom'
  )

plt

df2 = prepare.df.by.category('device_category')

plt = df2 %>% ggplot(aes(x=value, y=completion.time.in.minutes, color=device_category)) +
  geom_point(alpha = 0.5) +
  facet_wrap(vars(parameter), nrow=1, scales = "free_x") +
  xlab('') + ylab('completion time (in minute)') + 
  ggtitle('plot by device type') + 
  theme(
    strip.text = element_text(size = 5),
    axis.title.y = element_text(size = 8),
    legend.position = 'bottom'
  )

plt

df2 = prepare.df.by.category('assignment_type')

plt = df2 %>% ggplot(aes(x=value, y=completion.time.in.minutes, color=assignment_type)) +
  geom_point(alpha = 0.5) +
  facet_wrap(vars(parameter), nrow=1, scales = "free_x") +
  xlab('') + ylab('completion time (in minute)') + 
  ggtitle('plot by assignment type') + 
  theme(
    strip.text = element_text(size = 5),
    axis.title.y = element_text(size = 8),
    legend.position = 'bottom'
  )

plt

df2 = prepare.df.by.category('project_instruction_language')

plt = df2 %>% ggplot(aes(x=value, y=completion.time.in.minutes, color=project_instruction_language)) +
  geom_point(alpha = 0.5) +
  facet_wrap(vars(parameter), nrow=1, scales = "free_x") +
  xlab('') + ylab('completion time (in minute)') + 
  ggtitle('plot by instruction language') + 
  theme(
    strip.text = element_text(size = 5),
    axis.title.y = element_text(size = 8),
    legend.position = 'bottom'
  )

plt

Completion time by project attribute (has_audio, has_button, has_checkbox_input, has_externalHtml, has_fileAudio_input, has_fileImg_input, has_fileVideo_input, has_file_input, has_iframe, has_image, has_radio_input, has_sbs, has_select_input, has_sourcesRecorder_input, has_string_input, has_suggest_input, has_textarea_input, has_video)

Each plot is categorized by assignment_status, device_category, assignment_type

prepare.df.by.category = function(cat){
  
  data.by.cat = df %>% select(project_has_audio, project_has_externalHtml, project_has_fileAudio_input, project_has_fileImg_input, project_has_file_input, project_has_iframe, project_has_sbs, project_has_select_input, project_has_sourcesRecorder_input, project_has_suggest_input, project_has_textarea_input, project_has_video, cat, completion.time.in.minutes)
  
  data.by.cat = data.by.cat %>% 
  pivot_longer(cols=c('project_has_audio', 'project_has_externalHtml', 'project_has_fileAudio_input', 'project_has_fileImg_input', 'project_has_file_input', 'project_has_iframe', 'project_has_sbs', 'project_has_select_input', 'project_has_sourcesRecorder_input', 'project_has_suggest_input', 'project_has_textarea_input', 'project_has_video'), names_to = 'parameter')
  
  return(data.by.cat)
}

df2 = prepare.df.by.category('assignment_status')

plt = df2 %>% ggplot(aes(x=factor(value), y=completion.time.in.minutes, color=assignment_status)) +
  geom_boxplot() +
  facet_wrap(vars(parameter), nrow=4, scales = "free_x") +
  xlab('') + ylab('completion time (in minute)') + 
  ggtitle('plot by assignment status') + 
  theme(
    strip.text = element_text(size = 12),
    axis.text.x = element_text(size=15),
    axis.text.y = element_text(size = 15),
    legend.position = 'bottom'
  )

plt

df2 = prepare.df.by.category('device_category')

plt = df2 %>% ggplot(aes(x=factor(value), y=completion.time.in.minutes, color=device_category)) +
  geom_boxplot() +
  facet_wrap(vars(parameter), nrow=4, scales = "free_x") +
  xlab('') + ylab('completion time (in minute)') + 
  ggtitle('plot by assignment status') + 
  theme(
    strip.text = element_text(size = 12),
    axis.text.x = element_text(size=15),
    axis.text.y = element_text(size = 15),
    legend.position = 'bottom'
  )

plt

df2 = prepare.df.by.category('assignment_type')

plt = df2 %>% ggplot(aes(x=factor(value), y=completion.time.in.minutes, color=assignment_type)) +
  geom_boxplot() +
  facet_wrap(vars(parameter), nrow=4, scales = "free_x") +
  xlab('') + ylab('completion time (in minute)') + 
  ggtitle('plot by assignment status') + 
  theme(
    strip.text = element_text(size = 12),
    axis.text.x = element_text(size=15),
    axis.text.y = element_text(size = 15),
    legend.position = 'bottom'
  )

plt

df2 = prepare.df.by.category('project_instruction_language')

Note: Using an external vector in selections is ambiguous.
ℹ Use `all_of(cat)` instead of `cat` to silence this message.
ℹ See <https://tidyselect.r-lib.org/reference/faq-external-vector.html>.
This message is displayed once per session.

plt = df2 %>% ggplot(aes(x=factor(value), y=completion.time.in.minutes, color=project_instruction_language)) +
  geom_boxplot() +
  facet_wrap(vars(parameter), nrow=4, scales = "free_x") +
  xlab('') + ylab('completion time (in minute)') + 
  ggtitle('plot by assignment status') + 
  theme(
    strip.text = element_text(size = 12),
    axis.text.x = element_text(size=15),
    axis.text.y = element_text(size = 15),
    legend.position = 'bottom'
  )

plt

Completion time by project attribute (instruction_FK, instruction_language, instruction_len, instruction_wordCount, required_fields, spec_length)

Each plot is categorized by assignment_status, device_category, assignment_type, instruction_language

prepare.df.by.category = function(cat){
  
  # data.by.cat = df %>% select(project_instruction_FK, project_instruction_len, project_instruction_wordCount, project_required_fields, project_spec_length, project_has_button,  project_has_checkbox_input, project_has_image, project_has_radio_input, project_has_string_input, completion.time.in.minutes, cat)
  data.by.cat = df %>% select(project_instruction_FK, project_instruction_len, project_instruction_wordCount, project_required_fields, project_spec_length, completion.time.in.minutes, cat)
  
  data.by.cat = data.by.cat %>% 
  # pivot_longer(cols=c('project_instruction_FK', 'project_instruction_len', 'project_instruction_wordCount', 'project_required_fields', 'project_spec_length', 'project_has_button',  'project_has_checkbox_input', 'project_has_image', 'project_has_radio_input', 'project_has_string_input'), names_to = 'parameter')
    pivot_longer(cols=c('project_instruction_FK', 'project_instruction_len', 'project_instruction_wordCount', 'project_required_fields', 'project_spec_length'), names_to = 'parameter')
  
  return(data.by.cat)
}

df2 = prepare.df.by.category('assignment_status')

plt = df2 %>% ggplot(aes(x=value, y=completion.time.in.minutes, color=assignment_status)) +
  geom_point(alpha = 0.5) +
  facet_wrap(vars(parameter), nrow=1, scales = "free_x") +
  xlab('') + ylab('completion time (in minute)') + 
  ggtitle('plot by assignment status') + 
  theme(
    strip.text = element_text(size = 8),
    axis.text.x = element_text(size=8, angle = 30),
    axis.text.y = element_text(size = 8),
    legend.position = 'bottom'
  )

plt

df2 = prepare.df.by.category('device_category')

plt = df2 %>% ggplot(aes(x=value, y=completion.time.in.minutes, color=device_category)) +
  geom_point(alpha = 0.5) +
  facet_wrap(vars(parameter), nrow=1, scales = "free_x") +
  xlab('') + ylab('completion time (in minute)') + 
  ggtitle('plot by device type') + 
  theme(
    strip.text = element_text(size = 8),
    axis.text.x = element_text(size=8, angle = 30),
    axis.text.y = element_text(size = 8),
    legend.position = 'bottom'
  )

plt

df2 = prepare.df.by.category('assignment_type')

plt = df2 %>% ggplot(aes(x=value, y=completion.time.in.minutes, color=assignment_type)) +
  geom_point(alpha = 0.5) +
  facet_wrap(vars(parameter), nrow=1, scales = "free_x") +
  xlab('') + ylab('completion time (in minute)') + 
  ggtitle('plot by assignment type') + 
  theme(
    strip.text = element_text(size = 8),
    axis.text.x = element_text(size=8, angle = 30),
    axis.text.y = element_text(size = 8),
    legend.position = 'bottom'
  )

plt

df2 = prepare.df.by.category('project_instruction_language')

plt = df2 %>% ggplot(aes(x=value, y=completion.time.in.minutes, color=project_instruction_language)) +
  geom_point(alpha = 0.5) +
  facet_wrap(vars(parameter), nrow=1, scales = "free_x") +
  xlab('') + ylab('completion time (in minute)') + 
  ggtitle('plot by instruction language') + 
  theme(
    strip.text = element_text(size = 8),
    axis.text.x = element_text(size=8, angle = 30),
    axis.text.y = element_text(size = 8),
    legend.position = 'bottom'
  )

plt

Completion time by project attribute (instruction_FK, instruction_language, instruction_len, instruction_wordCount, required_fields, spec_length, project_has_button, project_has_checkbox_input, project_has_image, project_has_radio_input, project_has_string_input)

Each plot is categorized by assignment_status, device_category, assignment_type, instruction_language

prepare.df.by.category = function(cat){
  
  # data.by.cat = df %>% select(project_instruction_FK, project_instruction_len, project_instruction_wordCount, project_required_fields, project_spec_length, project_has_button,  project_has_checkbox_input, project_has_image, project_has_radio_input, project_has_string_input, completion.time.in.minutes, cat)
  data.by.cat = df %>% select(project_has_button,  project_has_checkbox_input, project_has_image, project_has_radio_input, project_has_string_input, completion.time.in.minutes, cat)
  
  data.by.cat = data.by.cat %>% 
  # pivot_longer(cols=c('project_instruction_FK', 'project_instruction_len', 'project_instruction_wordCount', 'project_required_fields', 'project_spec_length', 'project_has_button',  'project_has_checkbox_input', 'project_has_image', 'project_has_radio_input', 'project_has_string_input'), names_to = 'parameter')
    pivot_longer(cols=c('project_has_button',  'project_has_checkbox_input', 'project_has_image', 'project_has_radio_input', 'project_has_string_input'), names_to = 'parameter')
  
  return(data.by.cat)
}

df2 = prepare.df.by.category('assignment_status')

plt = df2 %>% ggplot(aes(x=value, y=completion.time.in.minutes, color=assignment_status)) +
  geom_point(alpha = 0.5) +
  facet_wrap(vars(parameter), nrow=1, scales = "free_x") +
  xlab('') + ylab('completion time (in minute)') + 
  ggtitle('plot by assignment status') + 
  theme(
    strip.text = element_text(size = 10),
    axis.text.x = element_text(size=8, angle = 30),
    axis.text.y = element_text(size = 8),
    legend.position = 'bottom'
  )

plt

df2 = prepare.df.by.category('device_category')

plt = df2 %>% ggplot(aes(x=value, y=completion.time.in.minutes, color=device_category)) +
  geom_point(alpha = 0.5) +
  facet_wrap(vars(parameter), nrow=1, scales = "free_x") +
  xlab('') + ylab('completion time (in minute)') + 
  ggtitle('plot by device type') + 
  theme(
    strip.text = element_text(size = 8),
    axis.text.x = element_text(size=8, angle = 30),
    axis.text.y = element_text(size = 8),
    legend.position = 'bottom'
  )

plt

df2 = prepare.df.by.category('assignment_type')

plt = df2 %>% ggplot(aes(x=value, y=completion.time.in.minutes, color=assignment_type)) +
  geom_point(alpha = 0.5) +
  facet_wrap(vars(parameter), nrow=1, scales = "free_x") +
  xlab('') + ylab('completion time (in minute)') + 
  ggtitle('plot by assignment type') + 
  theme(
    strip.text = element_text(size = 8),
    axis.text.x = element_text(size=8, angle = 30),
    axis.text.y = element_text(size = 8),
    legend.position = 'bottom'
  )

plt

df2 = prepare.df.by.category('project_instruction_language')

plt = df2 %>% ggplot(aes(x=value, y=completion.time.in.minutes, color=project_instruction_language)) +
  geom_point(alpha = 0.5) +
  facet_wrap(vars(parameter), nrow=1, scales = "free_x") +
  xlab('') + ylab('completion time (in minute)') + 
  ggtitle('plot by instruction language') + 
  theme(
    strip.text = element_text(size = 8),
    axis.text.x = element_text(size=8, angle = 30),
    axis.text.y = element_text(size = 8),
    legend.position = 'bottom'
  )

plt